(1) Field of the Invention
The present invention relates to the preparation of sorbents, which are particularly useful for the removal of sulfur dioxide and sulfur trioxide from gas mixtures, particularly from the flue gases of coal-burning power plants.
(2) Prior Art
In fossil-fuel-fired power plants, the sulfur content of the feed coal is oxidized during combustion to sulfur oxides (SO.sub.2 and SO.sub.3, commonly referred to as "SO.sub.x "), which are released through stacks to the atmosphere, and are responsible for deposition as "acid rain". Analyses of flue gas produced by power plants burning coal before desulfurization, show 0.5%-0.2% SO.sub.2 and about 0.005% SO.sub.3. Control of SO.sub.x emission is mandated by the U.S. Environmental Protection Agency (EPA), and various studies are under way to develop methods for its removal from flue gas streams.
Formation of SO.sub.x in combustion processes can be reduced by modifying the burner design and combustion system, by changing the operating conditions and by using fuels with lower sulfur contents. The most popular and inexpensive method of reducing SO.sub.x emission is the addition of reactive dry sorbents with the fuel. Accordingly at present, SO.sub.x removal is most often accomplished by using lime (CaO) or lime stone (CaCO.sub.3). Several other basic sorbents like MgO, ZnO also are found to be effective in removing SO.sub.x. For a review on dry sorbents see for example, Komppa, V., "Dry Adsorption Processes for Removal of SO.sub.x and NO.sub.x in Flue Gases--a review," Paperii ja Puu, 5, 401 to 405 (1986).
Use of Group 2 (formerly Group IIA) metal oxides such as magnesium and calcium oxides as SO.sub.x sorbents has been disclosed in several patent disclosures and recent examples include U.S. Pat. Nos. 3,835,031 and 3,699,037. Several other metal oxides of varying effectiveness as SO.sub.x sorbents are described in U.S. Pat. No. 4,153,534 which include oxides such as sodium, scandium, titanium, iron, chromium, molybdenum, manganese, cobalt, nickel, copper, zinc, cadmium, rare earth metals, and lead.
In typical coal-fired power plants the ground sorbent, for example lime or limestone, is added into boilers along with coal or sprayed into towers as a slurry to contact the flue gas. The SO.sub.2 reacts with calcium hydroxide to form a calcium sulfite slurry which is then partially oxidized with air to calcium sulfate. In this way the sulfur oxides are retained as harmless solid compounds which can be removed from the stack gas by electrostatic precipitation or other standard methods. Such a process is potentially attractive for retrofitting existing power plants since no major structural alterations are required.
A major problem with this type of process is low utilization of the oxide sorbents. The rate of adsorption of SO.sub.x declines rapidly with increasing conversion, due to mass transfer limitation and low reactivity of SO.sub.2. Hence in the relatively short contact time available, only a small fraction of the sorbent reacts. In principle the problem of low utilization of the sorbents may be solved by reducing the particle size, but in practice, the particle size required for a reasonable level of utilization may be too small to achieve economically by conventional grinding or fragmentation methods.
Thermodynamic calculations indicate that the capture of sulfur trioxide with metal oxides is more favorable compared to sulfur dioxide. Several experimental results have suggested that catalytic oxidation of sulfur dioxide to sulfur trioxide can be beneficial for stack gas desulfurization. Kocheffe & Karman in Cand. J. Chem. Eng., 63, 971 to 977 (1985) has shown that the rate of reaction of SO.sub.3 with Ca, Mg and ZnO is greater than that of sulfur dioxide with the same oxides under identical conditions. Furthermore, inclusion of Fe.sub.2 O.sub.3 (as a SO.sub.2 oxidation catalyst) leads to more effective utilization of the lime. The addition of a small amount of Fe.sub.2 O.sub.3 gave both a more rapid initial uptake rate and a much higher final conversion of the lime (80-90%). In the absence of an oxidation catalyst the rate of SO.sub.2 absorption declined sharply at about 70% conversion.
A similar approach has been employed in designing SO.sub.x sorbents for fluid catalytic cracking (FCC) processing of petroleum. These sorbents, among other things, are mostly alkaline earth metal spinels containing one or more other metal components capable of oxidizing sulfur dioxide. For example, U.S. Pat. Nos. 4,472,532 and 4,492,678 relate to the incorporation of iron, chromium, vanadium, manganese, gallium, boron, cobalt, platinum, cerium as oxidation catalyst.
Therefore, in designing improved sorbents for SO.sub.x removal, one must synthesize materials that will (i) oxidize SO.sub.2 to SO.sub.3, (ii) chemisorb the SO.sub.3 formed, and (iii) be able to release the adsorbed SO.sub.x for the regeneration of the sorbents or form stable materials for the safe deposition of the spent solid sorbents. The SO.sub.x emitted from these spent sorbents can be captured safely and can be utilized in sulfuric acid or sulfur production. These requirements can be achieved by employing high surface area basic sorbents which contain oxidation catalysts to facilitate the conversion of SO.sub.2 to SO.sub.3. This invention discloses a new class of SO.sub.x sorbents prepared from Layered Double Hydroxides (LDHs).
European Patent Application EP-A 278 535 has recently, described a catalyst composition suitable for the refining of heavy sulfur- and metal-containing petroleum feeds. Thus, the catalyst composition according to the disclosure contained a catalytically active zeolitic material such as ZSM-5, ZSM-11 etc. for the conversion of hydrocarbons, an anionic clay material with an LDH structure for the binding and removal of sulfur oxides, and a matrix material such as kaolin or alumina. Preferred catalyst compositions contained 1 to 30 percent amounts of anionic clay compositions, based on total catalyst composition.
In this present invention by us, a focus has been made to develop sorbent compositions suitable for diminishing SO.sub.x from flue gas streams particularly from coal-fired power plants. Emphasis was made in this invention to develop sorbent compositions which give better SO.sub.x uptake in shorter time duration to overcome the low utilization of common oxide sorbents such as CaO and MgO due to mass transfer limitation and low reactivity of SO.sub.2. These objectives and others will become apparent in the following discussions and furthermore, the superiority of the sorbents disclosed in this invention compared to compositions according to EP-A 278 535 will become clear.
LDHs are a group of anionic clay minerals. These have positively charged sheets of metal hydroxides, between which are located anions and some water molecules. Most common LDHs are based on double hydroxides of such main group metals as Mg, and Al and transition metals such as Ni, Co, Cr, Zn and Fe etc. These clays have structure similar to brucite [Mg(OH).sub.2 ] in which the magnesium ions are octahedrally surrounded by hydroxyl groups with the resulting octahedra sharing edges to form infinite sheets. In the LDHs, some of the magnesium is isomorphously replaced by a trivalent ion, such as Al.sup.3+. The Mg.sup.2+, Al.sup.3+, OH.sup.- layers are then positively charged, necessitating charge balancing by insertion of anions between the layers.
One such anionic clay is hydrotalcite in which the carbonate ion is the interstitial anion, and has the idealized unit cell formula [Mg.sub.6 Al.sub.2 (OH).sub.16 ](CO.sub.3).4H.sub.2 O. However the ratio of Mg/Al in hydrotalcite-like can vary between 1.7 and 4 and various other divalent and trivalent ions may be substituted for Mg and Al. In addition, the anion which is carbonate in hydrotalcite, can be varied in synthesis by a large number of simple anions such as NO.sub.3.sup.-, Cl.sup.-, OH.sup.-, SO.sub.4.sup.2- etc. These LDHs, based on their structure, fall into the Pyroaurite-Sjogrenite group, where brucite-like layers carrying a net positive charge alternate with layers in which the oxygen atoms of carbonate groups and water molecules are distributed on a single set of sites.
Hydrocalumite and related synthetic compounds also have a layered structure in which positively charged metal hydroxide layers alternate with the interlayers containing anions and water. The hydroxide layers contain specific combinations of metal ions derived from on one hand divalent calcium cations and on the other from trivalent cations of metals such as iron, or more particularly, aluminium. The interlayers contain anions such as OH.sup.-, SO.sub.4.sup.2-, Cl.sup.-, NO.sub.3.sup.- and, in particular CO.sub.3.sup.2-. The general formula for the group is [Ca.sub.2 M.sup.3+ (OH).sub.6 ]X.yH.sub.2 O, where M.sup.3+ is a tripositive ion and typically Al.sup.3+, X is a singly charged anion or equal amounts of more highly charged ones, and y is between 2 and 6. As in the Pyroaurite-Sjogrenite group, principal layers alternate with inter-layers, the principal layers having the composition [Ca.sub.2 M.sup.3+ (OH).sub.6 ].sup.+ and the interlayers consisting of water molecules and anion X. However, because of the difference in size between the Ca.sup.2+ and Al.sup.3+ ions, the M.sup.2+ :M.sup.3+ ratio is fixed at 2:1 and their arrangement is ordered. The only known natural mineral in the group is hydrocalumite the composition of which is approximately [Ca.sub.2 Al(OH).sub.6 ](OH).sub.0.75 (CO.sub.3).sub.0.125.2.5H.sub.2 O, but there are many synthetic analogues such as [Ca.sub.2 Fe(OH).sub.6 ](SO.sub.4).sub.0.5.3H.sub.2 O, [Ca.sub.2 Al(OH).sub.6 ](OH).6H.sub.2 O etc.
The syntheses of LDHs are generally simple, and the so called "precipitation method" is most popular. If a carbonate-containing product is desired, then the aqueous solution of magnesium and aluminum salts, i.e., nitrate, or chloride, is added to a aqueous solution of sodium hydroxide-carbonate with good mixing at room temperature. The resulting amorphous precipitate is then heated for several hours at 60.degree.-200.degree. C. to obtain a crystalline material. Washing and drying complete the synthesis in quantitative yield. By employing this precipitation method, replacement of all or part of Mg.sup.2+ with M.sup.II ions such as Ca.sup.2+, Zn.sup.2+, Cu.sup.2+ etc., or replacement of Al.sup.3+ with other M.sup.III ions such as Fe.sup.3+, Cr.sup.3+ etc., is also possible.
Another important aspect of the synthesis of these materials is the variation of the nature of the interstitial anion. The preparation of hydrotalcite-like materials with anions other than carbonate in pure form requires special procedures, because LDH incorporates carbonate in preference to other anions. Most of the time the smaller anions are introduced to the LDH structure, via the precipitation method by using the desired anion solutions instead of carbonate. However, in these methods the synthesis has to be carried out in an anaerobic condition to prevent carbonate contamination from the atmospheric carbon dioxide. These methods of preparation of LDHs have been described in prior art publications, particular reference being made to following review journal articles by S. L. Suib et. al., in Solid State Ionics, 26, 77 to 86 (1988), and W. T. Reichel in CHEMTECH, 58 to 63 (1986).
Process for the synthesis of hydrotalcite-like clays also have been the subject of a number of patents. Miyata et. al. in U.S. Pat. Nos. 3,796,792, 3,879,523, and 3,879,525 describes hydrotalcite-like derivatives with both cationic layer and anionic substitution including the smaller transition metal anions like CrO.sub.4.sup.2-, MoO.sub.4.sup.2-, and Mo.sub.2 O.sub.7.sup.2-. Both composition and preparative methods are described, and the compositions are said to be useful for catalytic purposes, absorbents, desiccants and the like. Synthetic hydrotalcite-like derivatives with small anions, including anions of transition elements, and also large organic anions such as long chain aliphatic dicarboxylates, are shown to catalyze aldol condensation effectively.
The nature of the thermal decomposition of LDHs especially the hydrotalcite-like materials, have been studied in detail. For example, upon thermolysis, hydrotalcite [Mg.sub.6 Al.sub.2 (OH).sub.16 ](CO.sub.3).4H.sub.2 O loses weight in two stages. First, it loses the four interstitial water molecules when heated to 200.degree. C., while retaining the skeletal hydroxide and the interlayer carbonate. Additional heating from 275.degree. C. to 450.degree. C. results in the simultaneous loss of hydroxyl groups and carbonate as water and carbon dioxide, respectively. Reichle in J. Catal. 101, 352 to 359 (1986) has shown that this heating of hydrotalcite was accompanied by an increase in the surface area from about 120 to about 230 m.sup.2 /g (N.sub.2 /BET) and a doubling of pore volume (0.6 to 1.0 cm.sup.3 /g, Hg intrusion). Further heating to higher temperatures causes lowering of surface area as well as reactivity. At 1000.degree. C., the formation of MgO and the spinel phase, MgAl.sub.2 O.sub.4 has been observed.
Thus, the thermal decomposition of LDHs leads to the formation of active metal oxides with fairly high basic character (pKa .ltoreq.35) and high surface area. These thermally treated materials should have exceptionally well-dispersed reactive metal centers, as judged from their catalytic properties. These properties have led us to synthesize and to utilize new thermally treated LDH materials as efficient sorbents for flue gas desulfurization.